Skip to content

merge development to master#144

Open
andreyorst wants to merge 57 commits into
masterfrom
development
Open

merge development to master#144
andreyorst wants to merge 57 commits into
masterfrom
development

Conversation

@andreyorst

Copy link
Copy Markdown
Collaborator

No description provided.

Panthevm and others added 30 commits March 12, 2026 16:34
- Add SQL autocompletion with table/function metadata fetching
- Add Tab to accept completion, Enter to suppress when completion active
- Improve line numbers styling and gutter active line colors
- Style completion matched text with link color and bold
- Add table icon for table completions
- Use CSS variables for search match colors
Changed line-height from --font-leading-3 (12px) to --font-leading-4 (16px)
for .typo-body-xs class. The previous value caused descender characters
(g, y, p, q) to be clipped when combined with overflow: hidden.
Color GET/POST/PUT/PATCH/DELETE in raw HTTP editor using
utility color tokens matching the request method selector.
- New getStructureDefinition prop on CodeEditor
- Autocomplete suggests fields based on StructureDefinition differential
- Resolves base definitions recursively (Resource, DomainResource, Element)
- Expands choice types to FHIR format (deceasedBoolean, deceasedDateTime)
- Adds primitive extensions (_birthDate, _active, etc.)
- Apply inserts typed snippets ("name": [{}], "active": , "gender": "")
- Caches loaded StructureDefinitions
isJsonValuePosition incorrectly matched "active": true, | as a value
position because [^"]* captured "true, ". Added early return when
beforeCursor ends with comma — the value is complete, not in progress.
Fixes primitive type completions (use FHIR names) and filters out
profiles from resource type completions in ofType()/is/as context.
Set gutter backgroundColor to --color-bg-primary instead of transparent
so that editor content does not show through line numbers when scrolling
horizontally.
Co-authored-by: Andrey Listopadov <andreyorst@gmail.com>
andreyorst and others added 27 commits April 8, 2026 17:22
Bumps [pnpm/action-setup](https://github.com/pnpm/action-setup) from 4 to 6.
- [Release notes](https://github.com/pnpm/action-setup/releases)
- [Commits](pnpm/action-setup@v4...v6)

---
updated-dependencies:
- dependency-name: pnpm/action-setup
  dependency-version: '6'
  dependency-type: direct:production
  update-type: version-update:semver-major
...

Signed-off-by: dependabot[bot] <support@github.com>
Bumps [actions/upload-pages-artifact](https://github.com/actions/upload-pages-artifact) from 4 to 5.
- [Release notes](https://github.com/actions/upload-pages-artifact/releases)
- [Commits](actions/upload-pages-artifact@v4...v5)

---
updated-dependencies:
- dependency-name: actions/upload-pages-artifact
  dependency-version: '5'
  dependency-type: direct:production
  update-type: version-update:semver-major
...

Signed-off-by: dependabot[bot] <support@github.com>
The "none" variant rendered with bg-sidebar (grey) and no border,
while the "icon"/"offcanvas" variants render with bg-bg-primary
(white) and side-aware border. Inconsistent visual when using
Sidebar standalone (e.g. nested inside another page).

- Switch background to bg-bg-primary
- Add border-r/border-l based on `side` for `variant="sidebar"`
- Expose `data-side` and `data-variant` on the root div for consistency
Previously inputRef.current?.focus() was called synchronously inside
changeOpen, which fires before Radix mounts the popover content. Move
the focus into a useEffect keyed on open with a requestAnimationFrame
so the input is actually focused once it exists in the DOM.
The DnD feature in headless-tree sets `draggable=true` on every TreeItem,
which blocks text selection in nested inputs even when no reordering is
configured. Pull it (and the drop-line element) out of the feature list
when canReorder is false so consumers that don't enable reordering get
back native text selection inside their custom item views.
Renders issue.diagnostics in its own typo-body text-text-secondary
column (flex-1 + truncate) when present. expression-span is now
conditional so issues without expression don't render an empty cell.
Aidbox returns only the first 100 entries by default for
/fhir/StructureDefinition?kind=resource (and there is no link.next),
so getResourceTypes() may not see e.g. Patient/Resource — and
completion-provider then hides functions constrained to FHIR Resource
(notably getResourceKey, getReferenceKey).

Pass _count=1000 to widen the page until proper pagination is added.
The horizontalScroll handler translated wheel deltaY into scrollTo on
the TabsList, competing with the browser's native overflow-x scroll
when a trackpad swipe emits deltaX with small/negative deltaY noise —
yanking the list back during smooth scrolling.

Skip the handler for trackpad-shape events (deltaX !== 0 or
|deltaY| < 5). Mouse-wheel input (large deltaY, zero deltaX) still
translates vertical input into horizontal scroll.

Also drop the ResizeObserver-driven scrollIntoView in TabsBrowserList:
mid-animation observer fires reset scrollLeft to the active tab.
onTabChange still reveals newly added tabs with {block:'nearest',
inline:'nearest'} so it's a no-op when already visible.
Radix Select Viewport hardcodes h-(--radix-select-trigger-height) for the
popper position. Safari honors it strictly and clips the embedded Command
(search + list) to ~30px; Chromium/Firefox let the flex child overflow, so
the bug is invisible there. Override the viewport height to auto only for
Combobox.
Add paddingRight: 400px to .cm-content so the area under the floating
top-right menubar (Copy / Format / Mode toggle in REST Console etc.)
stays empty and horizontal scrolling reaches everything without
hiding behind the menubar.

Give .cm-gutters the editor background (was transparent, so scrolled
text was visible through the line numbers) and a 4px var(--color-bg-primary)
shadow that fades content sliding under the gutter — makes horizontal
scroll discoverable.
…ssion cookie

When a request carries an explicit Authorization header (e.g. from a REST
notebook cell), treat it as the sole credential: don't attach the session
cookie (credentials: omit) and don't redirect to login on 401, so the caller
sees the real response.
A gated instance (unactivated license, or an SSO login gate) answers navigation with a 302 to an HTML page instead of 401. fetch follows the redirect, so the UI never saw a 401 and did not redirect. Now a redirected response on a request without an explicit Authorization header navigates the browser to the final URL, surfacing the login/activation page.
Bumps [actions/checkout](https://github.com/actions/checkout) from 6 to 7.
- [Release notes](https://github.com/actions/checkout/releases)
- [Changelog](https://github.com/actions/checkout/blob/main/CHANGELOG.md)
- [Commits](actions/checkout@v6...v7)

---
updated-dependencies:
- dependency-name: actions/checkout
  dependency-version: '7'
  dependency-type: direct:production
  update-type: version-update:semver-major
...

Signed-off-by: dependabot[bot] <support@github.com>
Bumps [actions/setup-node](https://github.com/actions/setup-node) from 6 to 7.
- [Release notes](https://github.com/actions/setup-node/releases)
- [Commits](actions/setup-node@v6...v7)

---
updated-dependencies:
- dependency-name: actions/setup-node
  dependency-version: '7'
  dependency-type: direct:production
  update-type: version-update:semver-major
...

Signed-off-by: dependabot[bot] <support@github.com>
# Conflicts:
#	package.json
#	pnpm-lock.yaml
The merge pinned brace-expansion to ^5.0.7 (from master's audit fix), but
5.x is ESM-only with named exports, breaking minimatch@9.0.9 which does a
default import (`import expand from 'brace-expansion'`) — e.g. typedoc's
`doc` script failed. The tree now has both minimatch@9 (needs 2.x) and
minimatch@10 (needs 5.x), so no single major works.

Replace the blanket pin with per-major security patches covering only the
vulnerable ranges (GHSA-v6h2-p8h4-qcjw), letting each consumer resolve its
compatible major: brace-expansion 2.1.2 for minimatch@9, 5.0.7 for minimatch@10.
@andreyorst
andreyorst requested a review from Panthevm July 22, 2026 09:15
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants